def __init__(self):
self.model = Sequential()
self.model.add(
Conv2D(32,
kernel_size=(3, 3),
activation='relu',
input_shape=(48, 48, 1)))
self.model.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))
self.model.add(MaxPooling2D(pool_size=(2, 2)))
self.model.add(Dropout(0.25))
self.model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
self.model.add(MaxPooling2D(pool_size=(2, 2)))
self.model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
self.model.add(MaxPooling2D(pool_size=(2, 2)))
self.model.add(Dropout(0.25))
self.model.add(Flatten())
self.model.add(Dense(1024, activation='relu'))
self.model.add(Dropout(0.5))
self.model.add(Dense(7, activation='softmax'))
print("model path: {}".format(
resource_path("Detection\Weight\model-epoch-30.h5")))
self.model.load_weights(
resource_path("Detection\Weight\model-epoch-30.h5"))
def client_recv(self, *args):
if self.detect_thread.is_alive():
try:
self.client_stream_socket = KSocketClient()
self.client_stream_socket.kconnect(socket.gethostname(),
self.stream_port)
str_encoded = self.client_stream_socket.kreceive()
nparr = np.fromstring(str_encoded, np.uint8)
if nparr.size != 0:
img_decoded = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
if self.status == 'started':
cv2.imwrite(resource_path('assets/v.jpg'), img_decoded)
self.source = 'assets/v.jpg'
self.reload()
if self.emotion_color is not None:
self.app.set_emotion_color(self.emotion_color)
else:
self.source = 'assets/video.jpg'
self.reload()
os.remove('assets/v.jpg')
except ConnectionRefusedError:
print('Hello ConnectionRefusedError')
else:
if self.status == 'started':
self.detect_thread = threading.Thread(
target=self.detect_from_camera, daemon=True)
self.detect_thread.start()
emotionDetector = EmotionDetector()
session_info = SessionInfo(None, None, None, None, None)
video_out = "temp_vid.mp4"
video_writer = cv2.VideoWriter(
video_out, cv2.VideoWriter_fourcc(*'mp4v'), fps, (width, height))
while True:
hasFace = False
# Find haar cascade to draw bounding box around face
ret, frame = cap.read()
if not ret:
break
frame = cv2.flip(frame, 1)
video_writer.write(cv2.resize(frame,(width, height),interpolation = cv2.INTER_CUBIC))
facecasc = cv2.CascadeClassifier(resource_path('Detection\haarcascade_frontalface_default.xml'))
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
faces = facecasc.detectMultiScale(gray,scaleFactor=1.3, minNeighbors=5)
frameInfo = FrameInfo(None, None, None)
for (x, y, w, h) in faces:
hasFace = True
cv2.rectangle(frame, (x, y-50), (x+w, y+h+10), (255, 0, 0), 2)
roi_gray = gray[y:y + h, x:x + w]
cropped_img = np.expand_dims(np.expand_dims(cv2.resize(roi_gray, (48, 48)), -1), 0)
maxindex = emotionDetector.detect_emotion(cropped_img)
cv2.putText(frame, emotion_dict[maxindex], (x+20, y-60), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2, cv2.LINE_AA)
streamHandler.add_frame(maxindex)
if hasFace is not True:
streamHandler.add_frame(7)
def build(self):
self.theme_cls.primary_palette = 'BlueGray'
layout = Builder.load_file(resource_path('main.kv'))
return layout
def detect_from_camera(self):
try:
server_stream_socket = socket.socket(socket.AF_INET,
socket.SOCK_STREAM)
server_stream_socket.bind((socket.gethostname(), self.stream_port))
server_stream_socket.listen()
# prevents openCL usage and unnecessary logging messages
cv2.ocl.setUseOpenCL(False)
# dictionary which assigns each label an emotion (alphabetical order)
emotion_dict = {
7: "No face detected",
0: "Angry",
1: "Disgusted",
2: "Fearful",
3: "Happy",
4: "Neutral",
5: "Sad",
6: "Surprised"
}
emotion_colors = {
7: "#000000",
0: "#FF005A",
1: "#33CC33",
2: "#9933FF",
3: "#FFCC00",
4: "#996600",
5: "#0099FF",
6: "#33CCCC"
}
cap = cv2.VideoCapture(0)
streamHandler = EmotionStreamHandler()
emotionDetector = EmotionDetector()
sessionInfo = SessionInfo(None, None, None, None)
while True:
(connection, address) = server_stream_socket.accept()
hasFace = False
# Find haar cascade to draw bounding box around face
ret, frame = cap.read()
if not ret:
break
frame = cv2.flip(frame, 1)
facecasc = cv2.CascadeClassifier(
resource_path(
'Detection\haarcascade_frontalface_default.xml'))
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
faces = facecasc.detectMultiScale(gray,
scaleFactor=1.3,
minNeighbors=5)
frameInfo = FrameInfo(None, None, None)
for (x, y, w, h) in faces:
hasFace = True
roi_gray = gray[y:y + h, x:x + w]
cropped_img = np.expand_dims(
np.expand_dims(cv2.resize(roi_gray, (48, 48)), -1), 0)
maxindex = emotionDetector.detectEmotion(cropped_img)
cv2.rectangle(frame, (x, y - 50), (x + w, y + h + 10),
(255, 0, 0), 2)
cv2.putText(frame, emotion_dict[maxindex],
(x + 20, y - 60), cv2.FONT_HERSHEY_SIMPLEX, 1,
(255, 255, 255), 2, cv2.LINE_AA)
streamHandler.addFrame(maxindex)
self.emotion_color = emotion_colors[maxindex]
if hasFace is not True:
streamHandler.addFrame(7)
self.emotion_color = emotion_colors[7]
if streamHandler.warning:
if not self.warning_dialog._window:
self.show_warning_dialog()
else:
if self.warning_dialog._window:
self.hide_warning_dialog()
img = cv2.resize(frame, (400, 300),
interpolation=cv2.INTER_CUBIC)
encode_param = [int(cv2.IMWRITE_JPEG_QUALITY), 90]
result, img_encoded = cv2.imencode('.jpg', img, encode_param)
data_encoded = np.array(img_encoded)
str_encoded = data_encoded.tostring()
connection.sendall(str_encoded)
connection.close()
if self.status == 'ended':
break
Clock.unschedule(self.client_recv)
sessionInfo = streamHandler.finish()
for i in range(0, len(sessionInfo.periods)):
print("===={}==== size: {}".format(
emotion_dict[i], len(sessionInfo.periods[i])))
for period in sessionInfo.periods[i]:
print(period.__dict__)
duration = int(
round((period.periodEnd - period.periodStart) * 1000))
sessionEvaluator = SessionEvaluator()
sessionEvaluator.evaluate(sessionInfo)
cap.release()
cv2.destroyAllWindows()
except:
pass